Wireless Access

We just replaced our AP-125's with AP-325's in one of our administrative buildings. Every since the replacement, a bunch of users are reporting issues with Wi-Fi connectivity. Their workstation does not disconnect, but showed "No Internet Access" with the yellow icon over the Wi-Fi bars. In the event viewer logs, it indicates a few errors including:

1. Intel(R) Dual Band Wireless-AC 8265 : Has determined that the network adapter is not functioning properly.
2. Intel(R) Dual Band Wireless-AC 8265 : Has encountered an internal error and has failed.
3. The network interface "Intel(R) Dual Band Wireless-AC 8265" has begun resetting. There will be a momentary disruption in network connectivity while the hardware resets. Reason: The network driver requested that it be reset. This network interface has reset 8 time(s) since it was last initialized.
4. 5007 - TX/CMD timeout (TfdQueue hanged)

In my controller it shows the client-trail with Unspecified Error as part of the Deauth section. When I check the mgmt frames to the AP from the client using "show ap remote debug mgmt-frames <ap-name> <client-mac> it shows a deauth frame sourcing from the client's MAC to the AP's BSSID.

Everything tells me the client is the issue, but it only started once we swapped to AP-325's and 802.11ac. I remember having some issues with the Intel 7265 and 802.11ac early on and I think I had to disable VHT for it to function properly. I would like to avoid this if possible, but don't know if it's a similar issue. 

We are running 6.4.4.16 code on our 7220 controller pair supporting this office. We are finishing up the AP-125 replacements and plan to go to 8.5.0.5 when we get rid of those old APs 

Anybody have similar issues or suggestions on what else we can try or look into?

By changing from 802.11n to 802.11ac a lot has been changed in your environment and can introduce some client bugs that where hidden in the past.

First rule of thumb is check you have the latest intel drivers on your clients, there are a lot issued around at google for this intel chipset.

Because you change your access points its also a good idea to take a close look at your RF settings if they are changed and if they are still meet the requirements.

Maybe you can do som troubleshooting layer by layer.

Phase 1:

Is the SSID discovered and are the 802.11 frames do authenticate and associated succeed?

(HomeLAB-MC01) [MDC] *#show ap remote debug mgmt-frames ap-name HomeLAB-AP01 count 50

Phase 2:

Does L2 authentication WPA/WPA2/802.1x succeed? If the user is not in the station-table of the corresponding controller Phase2 is failing.

(HomeLAB-MC01) [MDC] *#show station-table

### Enable User-debugging ###
(HomeLAB-MM01) [##:##:##:##:##:##] (config) #logging user-debug 44:85:00:9b:48:80 subcat all level debugging
(HomeLAB-MM01) ^[##:##:##:##:##:##] (config) #write mem
(HomeLAB-MM01) ^[##:##:##:##:##:##] (config) #mdconnect

(HomeLAB-MC01) [MDC] *#show log user-debug 100

Phase 3:

Does the client get an IP address?

Phase 4: 

What is the security policy that take place for the client?

Phase 5: 

What network access doest the client get?

(HomeLAB-MM01) [##:##:##:##:##:##] (config) #logging network subcat dhcp level debugging
(HomeLAB-MM01) ^[##:##:##:##:##:##] (config) #write mem
(HomeLAB-MM01) ^[##:##:##:##:##:##] (config) #mdconnect

(HomeLAB-MC01) [MDC] *#show log network 50

Otherthings you can check.

What channels, channel-width,power-settings are used?

(HomeLAB-MC01) [MDC] *#show ap bss-table

"show user mac" give you a lot of information of a connected client.

(HomeLAB-MC01) [MDC] *#show user mac ##:##:##:##:##:##

I'll go through each of those steps again, but all of those seemed fine, the drop I had was not happening during roaming or the initial association/authentication process, it's after the workstation in question was connected for about 5 hours actively passing traffic and then randomly indicated those NIC issues. 

Since I recall issues with 802.11ac and the 7260 series, I disabled VHT in the RF profile (which was the only main change in config between AP-125 and AP-325) to see if that stabilizes the connectivity for the users. 

If this gets rid of the issue, then I at least know what is causing it (rules out all other changes between the APs) and still then points to a previously unidentified bug with the workstations in my opinion.